Influence of Chronic Kidney Disease on Presentation, Treatment and Outcome in Patients with Coronary Artery Disease

Presentation, Treatment and Outcome in Patients with Coronary Artery Disease

Robert Edfors

al degree (Ph.D.) 2019Robert Edfors Influence of Chronic Kidney Disease on Presentation, Treatment and Outcome in Patients with Coronary Artery Disease

Department of Medicine, Division of Cardiology Karolinska Institutet, Huddinge, Stockholm, Sweden

INFLUENCE OF CHRONIC KIDNEY DISEASE ON PRESENTATION, TREATMENT AND OUTCOME IN PATIENTS WITH CORONARY ARTERY DISEASE

Robert Edfors

Department of Medicine, Division of Cardiology Karolinska Institutet, Huddinge, Stockholm, Sweden

INFLUENCE OF CHRONIC KIDNEY DISEASE ON PRESENTATION, TREATMENT AND OUTCOME IN PATIENTS WITH CORONARY ARTERY DISEASE

Robert Edfors

Cover illustration by Kerstin Edfors

All previously published papers were reproduced with permission from the publisher.

Published by Karolinska Institutet.

Printed by E-print AB 2019

© Robert Edfors, 2019 ISBN 978-91-7831-616-8

Cover illustration by Kerstin Edfors

All previously published papers were reproduced with permission from the publisher.

Published by Karolinska Institutet.

Printed by E-print AB 2019

© Robert Edfors, 2019 ISBN 978-91-7831-616-8

To my family To my family

“In your life expect some trouble, when you worry you make it double, don't worry, be happy.”

Bobby McFerry, 1988

“In your life expect some trouble, when you worry you make it double, don't worry, be happy.”

Bobby McFerry, 1988

Department of Medicine, Division of Cardiology, Karolinska Institutet, Huddinge, Stockholm, Sweden

INFLUENCE OF CHRONIC KIDNEY DISEASE ON PRESENTATION, TREATMENT AND OUTCOME IN

PATIENTS WITH CORONARY ARTERY DISEASE

THESIS FOR DOCTORAL DEGREE (Ph.D.) AKADEMISK AVHANDLING

som för avläggande av medicine doktorsexamen vid Karolinska Institutet offentligen försvaras i föreläsningssal Birke aulan F 52, Karolinska Universitetssjukhuset i

Huddinge, fredagen den 13 december 2019 kl 09.00 by

Robert Edfors

Principal Supervisor:

Professor Tomas Jernberg Karolinska Institutet

Department of Clinical Sciences Danderyd University Hospital

Co-supervisors:

Karolina Szummer, PhD Karolinska Institutet Department of Medicine Division of Cardiology Professor Stefan James Uppsala University

Department of Medical Sciences Division of Cardiology Marie Evans, PhD Karolinska Institutet

Department of Clinical Science Intervention and Technology Division of Renal Medicine

Opponent:

Professor Sten Dalby Kristensen Aarhus University

Department of Clinical Medicine Division of Cardiological Medicine B

Examination Board:

Associate Professor Jan-Erik Karlsson Linköping University

Department of Medical Health Sciences Division of Cardiovascular Medicine Professor Bengt Fellström

Uppsala University

Department of Medical Sciences Division of Renal Medicine Associate Professor Nawzad Saleh Karolinska Institutet

Department of Medicine Division of Cardiology

Department of Medicine, Division of Cardiology, Karolinska Institutet, Huddinge, Stockholm, Sweden

INFLUENCE OF CHRONIC KIDNEY DISEASE ON PRESENTATION, TREATMENT AND OUTCOME IN

PATIENTS WITH CORONARY ARTERY DISEASE

THESIS FOR DOCTORAL DEGREE (Ph.D.) AKADEMISK AVHANDLING

som för avläggande av medicine doktorsexamen vid Karolinska Institutet offentligen försvaras i föreläsningssal Birke aulan F 52, Karolinska Universitetssjukhuset i

Huddinge, fredagen den 13 december 2019 kl 09.00 by

Robert Edfors

Principal Supervisor:

Professor Tomas Jernberg Karolinska Institutet

Department of Clinical Sciences Danderyd University Hospital

Co-supervisors:

Karolina Szummer, PhD Karolinska Institutet Department of Medicine Division of Cardiology Professor Stefan James Uppsala University

Department of Medical Sciences Division of Cardiology Marie Evans, PhD Karolinska Institutet

Department of Clinical Science Intervention and Technology Division of Renal Medicine

Opponent:

Professor Sten Dalby Kristensen Aarhus University

Department of Clinical Medicine Division of Cardiological Medicine B

Examination Board:

Associate Professor Jan-Erik Karlsson Linköping University

Department of Medical Health Sciences Division of Cardiovascular Medicine Professor Bengt Fellström

Uppsala University

Department of Medical Sciences Division of Renal Medicine Associate Professor Nawzad Saleh Karolinska Institutet

Department of Medicine Division of Cardiology

SAMMANFATTNING

Bakgrund

Ungefär en tredjedel av alla patienter med akut hjärtinfarkt har njursvikt, vilket innebär en markant ökad risk för såväl komplikationer som nya hjärthändelser. Syftet med detta avhandlingsarbete var att undersöka vilken betydelse olika grader av njursvikt har för överlevnad och kliniska utfall, att utvärdera utfallen efter olika medicinska- och interventionella behandlingar och hur de skiljer sig åt hos patienter med njursvikt samt att utifrån blodprovsanalyser studera olika biomarkörers koppling till njursvikt och kardiovaskulära händelser.

Metoder och resultat

Studie I: Data från det svenska hjärtsjukvårdsregistret SWEDEHEART användes för att undersöka hur olika grader av njursvikt påverkar risken för död och kliniska utfall efter en kranskärlsröntgen p.g.a. stabil kranskärlssjukdom. Resultaten påvisade att trots justering för flertalet kliniska bakgrundsfaktorer, samsjuklighet, grad av ischemi och olika revaskulariseringmetoder, hade patienter med njursvikt en signifikant ökad risk för död, återinsjuknande i akut hjärtinfarkt och hjärtsvikt.

Studie II: Registerstudie där alla patienter med hjärtinfarkt som registrerats i SWEDEHEART och som erhållit blodförtunnande behandling med antingen clopidogrel eller ticagrelor studerades.

Samkörning av flera register utfördes. Jämfört med clopidogrel, var behandling med ticagrelor förknippat med lägre risk för död, återinläggning med hjärtinfarkt och stroke, dock på bekostnad av en ökad blödningsrisk, hos både njurfriska och patienter med måttlig njursvikt. Bland patienter med uttalad njursvikt observerades inte någon säker nytta, men däremot ökade risker för blödning.

Studie III: Alla patienter i SWEDEHEART som genomgått kranskärlsröntgen och erhållit en kranskärlsstent (läkemedelsbärande stent (n-DES) eller stent av metall (BMS)) inkluderades i denna studie, där risken för förnyad förträngning (restenos) och akut ocklusion (stenttrombos) i stenten studerades utifrån njurfunktion. I jämförelse med BMS, innebar stentinläggning med n-DES en lägre 1-årsrisk för både restenos och stenttrombos hos patienter med lätt eller måttlig njursvikt, medan ingen skillnad noterades hos patienter med uttalad njursvikt.

Studie IV: SWEDEHEART användes för både provinsamling och prospektiv långtidsuppföljning hos 1,098 patienter som vårdats med akut hjärtinfarkt. Blodprover sparades i SWEDEHEART- biobank och 175 olika biomarkörers uttryck analyserades hypotesfritt angående kopplingar till njursvikt och kliniska utfall, som registrerades i befolknings- och patientregistret. Resultaten påvisade att flertalet av de sex stycken biomarkörer som var starkast relaterade till njursvikt också var starkt relaterade till risken för att drabbas av död, hjärtinfarkt och hjärtsvikt.

Slutsats

Njursvikt är en stark och oberoende riskmarkör för överdödlighet och kardiovaskulära utfall hos patienter med kranskärlssjukdom. Andra okända faktorer än de vi kunnat justera för (riskfaktorer, samsjuklighet, CAD, behandlingar) har betydelse för den försämrade prognos hos patienter med kranskärlssjukdom och njursvikt. Hos njurfriska och patienter med måttligt sänkt njurfunktion, medförde blodförtunnande behandling med ticagrelor jämfört med clopidogrel samt valet av stenttyp;

ökad blödningsrisk, lägre risk för död, hjärtinfarkt och stroke, respektive lägre risk för stenthändelser, medan inga fördelar noterades hos patienter med uttalad njursvikt. Det påvisades att ett fåtal biomarkörer hade ett starkt gemensamt samband till både njursvikt, död, hjärtinfarkt och hjärtsvikt, vilket tyder på att biomarkörernas underliggande patologiska mekanismer kan förklara den försämrade långtidsprognos som ses hos patienter med kranskärlssjukdom och njursvikt.

SAMMANFATTNING

Bakgrund

Ungefär en tredjedel av alla patienter med akut hjärtinfarkt har njursvikt, vilket innebär en markant ökad risk för såväl komplikationer som nya hjärthändelser. Syftet med detta avhandlingsarbete var att undersöka vilken betydelse olika grader av njursvikt har för överlevnad och kliniska utfall, att utvärdera utfallen efter olika medicinska- och interventionella behandlingar och hur de skiljer sig åt hos patienter med njursvikt samt att utifrån blodprovsanalyser studera olika biomarkörers koppling till njursvikt och kardiovaskulära händelser.

Metoder och resultat

Studie I: Data från det svenska hjärtsjukvårdsregistret SWEDEHEART användes för att undersöka hur olika grader av njursvikt påverkar risken för död och kliniska utfall efter en kranskärlsröntgen p.g.a. stabil kranskärlssjukdom. Resultaten påvisade att trots justering för flertalet kliniska bakgrundsfaktorer, samsjuklighet, grad av ischemi och olika revaskulariseringmetoder, hade patienter med njursvikt en signifikant ökad risk för död, återinsjuknande i akut hjärtinfarkt och hjärtsvikt.

Studie II: Registerstudie där alla patienter med hjärtinfarkt som registrerats i SWEDEHEART och som erhållit blodförtunnande behandling med antingen clopidogrel eller ticagrelor studerades.

Samkörning av flera register utfördes. Jämfört med clopidogrel, var behandling med ticagrelor förknippat med lägre risk för död, återinläggning med hjärtinfarkt och stroke, dock på bekostnad av en ökad blödningsrisk, hos både njurfriska och patienter med måttlig njursvikt. Bland patienter med uttalad njursvikt observerades inte någon säker nytta, men däremot ökade risker för blödning.

Studie III: Alla patienter i SWEDEHEART som genomgått kranskärlsröntgen och erhållit en kranskärlsstent (läkemedelsbärande stent (n-DES) eller stent av metall (BMS)) inkluderades i denna studie, där risken för förnyad förträngning (restenos) och akut ocklusion (stenttrombos) i stenten studerades utifrån njurfunktion. I jämförelse med BMS, innebar stentinläggning med n-DES en lägre 1-årsrisk för både restenos och stenttrombos hos patienter med lätt eller måttlig njursvikt, medan ingen skillnad noterades hos patienter med uttalad njursvikt.

Studie IV: SWEDEHEART användes för både provinsamling och prospektiv långtidsuppföljning hos 1,098 patienter som vårdats med akut hjärtinfarkt. Blodprover sparades i SWEDEHEART- biobank och 175 olika biomarkörers uttryck analyserades hypotesfritt angående kopplingar till njursvikt och kliniska utfall, som registrerades i befolknings- och patientregistret. Resultaten påvisade att flertalet av de sex stycken biomarkörer som var starkast relaterade till njursvikt också var starkt relaterade till risken för att drabbas av död, hjärtinfarkt och hjärtsvikt.

Slutsats

Njursvikt är en stark och oberoende riskmarkör för överdödlighet och kardiovaskulära utfall hos patienter med kranskärlssjukdom. Andra okända faktorer än de vi kunnat justera för (riskfaktorer, samsjuklighet, CAD, behandlingar) har betydelse för den försämrade prognos hos patienter med kranskärlssjukdom och njursvikt. Hos njurfriska och patienter med måttligt sänkt njurfunktion, medförde blodförtunnande behandling med ticagrelor jämfört med clopidogrel samt valet av stenttyp;

ökad blödningsrisk, lägre risk för död, hjärtinfarkt och stroke, respektive lägre risk för stenthändelser, medan inga fördelar noterades hos patienter med uttalad njursvikt. Det påvisades att ett fåtal biomarkörer hade ett starkt gemensamt samband till både njursvikt, död, hjärtinfarkt och hjärtsvikt, vilket tyder på att biomarkörernas underliggande patologiska mekanismer kan förklara den försämrade långtidsprognos som ses hos patienter med kranskärlssjukdom och njursvikt.

ABSTRACT

Background

About one third of patients with myocardial infarction (MI) have renal dysfunction (RD).

Concomitant coronary artery disease (CAD) and RD is accompanied by a markedly higher risk of death and subsequent cardiovascular (CV) events. The aims of this thesis were to examine the association between degree of RD, mortality and subsequent CV events in patients with stable CAD, to evaluate outcomes of different medical- and interventional treatment regimens in relation to renal function in patients with MI or undergoing percutaneous coronary intervention (PCI), as well as to study various biomarkers and their associations to RD and long-term outcomes in patients with MI.

Methods and results

Study I: We used the Swedish Web-system for Enhancement and Development of Evidence-based care in Heart disease Evaluated According to Recommended Therapies (SWEDEHEART) registry to study the associations between renal function, death and CV outcomes in patients undergoing coronary angiography due to stable CAD. Despite adjusting for clinical background, risk factors, co- morbidities, severity of CAD and mode of revascularization, patients with RD had a significantly higher risk of death, subsequent MI readmission and heart failure hospitalization compared to patients without RD.

Study II: Follow-up data in MI survivors enrolled in the SWEDEHEART was used to study the association between ticagrelor versus clopidogrel and death, MI or stroke and the risk of bleeds.

Ticagrelor as compared with clopidogrel, was associated with lower risk for CVD outcomes and a higher bleeding risk in patients with normal- and moderately reduced renal function. In patients with severe RD, bleeds were more abundant in patients and the benefits less clear.

Study III: Observational SWEDEHEART study, that compared the 1-year risk of in-stent restenosis and stent thrombosis in patients treated with coronary artery stenting using either bare metal- (BMS) or newer generation drug eluting stents (n-DES), in relation to renal function. N-DES, as compared with BMS, was associated with a lower 1-year risk of in-stent restenosis and stent thrombosis in patients with normal- and moderately reduced renal function, whereas no differences were observed between stent type and stent events in patients with severe RD.

Study IV: SWEDEHEART was utilized for blood sample collection and prospective long-term follow up in 1,098 MI patients. Samples were saved in the SWEDEHEART-biobank and subsequently an untargeted analysis of 175 different biomarkers was conducted to study associations with RD, subsequent death, MI readmission and heart failure hospitalization. Six of the strongest biomarkers for RD also shared a strong variable importance for the studied long-term outcomes.

Conclusion

RD is a strong and independent marker of worse outcomes in patients with CAD. Other unknown factors that were not possible to adjust for may play an important role for the risk of adverse outcomes observed in CAD patients with RD. In patients with no- or moderate RD, ticagrelor versus clopidogrel and the choice of coronary stent type were associated with higher risk of bleeds, lower risk of death, stroke or MI, as well as lower risk of stent events, respectively. However, in patients with severe RD no beneficial effects were observed following the same treatment regimens. The identification of six biomarkers with strong mutual associations with both RD and CV outcomes, may indicate the underlying mechanisms that may contribute to the poor prognosis seen in patients with concomitant MI and RD.

ABSTRACT

Background

About one third of patients with myocardial infarction (MI) have renal dysfunction (RD).

Concomitant coronary artery disease (CAD) and RD is accompanied by a markedly higher risk of death and subsequent cardiovascular (CV) events. The aims of this thesis were to examine the association between degree of RD, mortality and subsequent CV events in patients with stable CAD, to evaluate outcomes of different medical- and interventional treatment regimens in relation to renal function in patients with MI or undergoing percutaneous coronary intervention (PCI), as well as to study various biomarkers and their associations to RD and long-term outcomes in patients with MI.

Methods and results

Study I: We used the Swedish Web-system for Enhancement and Development of Evidence-based care in Heart disease Evaluated According to Recommended Therapies (SWEDEHEART) registry to study the associations between renal function, death and CV outcomes in patients undergoing coronary angiography due to stable CAD. Despite adjusting for clinical background, risk factors, co- morbidities, severity of CAD and mode of revascularization, patients with RD had a significantly higher risk of death, subsequent MI readmission and heart failure hospitalization compared to patients without RD.

Study II: Follow-up data in MI survivors enrolled in the SWEDEHEART was used to study the association between ticagrelor versus clopidogrel and death, MI or stroke and the risk of bleeds.

Ticagrelor as compared with clopidogrel, was associated with lower risk for CVD outcomes and a higher bleeding risk in patients with normal- and moderately reduced renal function. In patients with severe RD, bleeds were more abundant in patients and the benefits less clear.

Study III: Observational SWEDEHEART study, that compared the 1-year risk of in-stent restenosis and stent thrombosis in patients treated with coronary artery stenting using either bare metal- (BMS) or newer generation drug eluting stents (n-DES), in relation to renal function. N-DES, as compared with BMS, was associated with a lower 1-year risk of in-stent restenosis and stent thrombosis in patients with normal- and moderately reduced renal function, whereas no differences were observed between stent type and stent events in patients with severe RD.

Study IV: SWEDEHEART was utilized for blood sample collection and prospective long-term follow up in 1,098 MI patients. Samples were saved in the SWEDEHEART-biobank and subsequently an untargeted analysis of 175 different biomarkers was conducted to study associations with RD, subsequent death, MI readmission and heart failure hospitalization. Six of the strongest biomarkers for RD also shared a strong variable importance for the studied long-term outcomes.

Conclusion

RD is a strong and independent marker of worse outcomes in patients with CAD. Other unknown factors that were not possible to adjust for may play an important role for the risk of adverse outcomes observed in CAD patients with RD. In patients with no- or moderate RD, ticagrelor versus clopidogrel and the choice of coronary stent type were associated with higher risk of bleeds, lower risk of death, stroke or MI, as well as lower risk of stent events, respectively. However, in patients with severe RD no beneficial effects were observed following the same treatment regimens. The identification of six biomarkers with strong mutual associations with both RD and CV outcomes, may indicate the underlying mechanisms that may contribute to the poor prognosis seen in patients with concomitant MI and RD.

LIST OF SCIENTIFIC PAPERS

I. Edfors R, Szummer K, Evans M, Carrero JJ, Spaak J, James SK, Lagerqvist B and Jernberg T. Renal function is associated with long-term outcomes independent of degree of atherosclerosis: 6-year data from the Swedish Coronary Angiography and Angioplasty Registry.

Eur Heart J Qual Care Clin Outcomes. 2016;2:91-98.

II. Edfors R, Sahlen A, Szummer K, Renlund H, Evans M, Carrero JJ, Spaak J, James SK, Lagerqvist B, Varenhorst C and Jernberg T. Outcomes in patients treated with ticagrelor versus clopidogrel after acute myocardial infarction stratified by renal function.

Heart. 2018;104:1575-1582.

III. Edfors R, James SK, Szummer K, Evans M, Carrero JJ, Faxén J, Persson J, Spaak J, Varenhorst C, Jernberg T, Lagerqvist B. SWEDEHEART-1-year data show no benefit of newer generation drug-eluting stents over bare-metal stents in patients with severe kidney dysfunction following percutaneous coronary intervention.

Coron Artery Dis. In press.

IV. Edfors R, Spaak J, Lindhagen L, Andell P, Baron B, Evans M, Erlinge D, James SK, Kahan T, Lindahl B, Rezeli M, Marko-Varga G, Mörtberg J, Salzinger B, Szummer K, Wallén H, Jernberg T. Use of proteomics to identify biomarkers associated with chronic kidney disease and long-term outcomes in patients with myocardial infarction.

Manuscript.

LIST OF SCIENTIFIC PAPERS

I. Edfors R, Szummer K, Evans M, Carrero JJ, Spaak J, James SK, Lagerqvist B and Jernberg T. Renal function is associated with long-term outcomes independent of degree of atherosclerosis: 6-year data from the Swedish Coronary Angiography and Angioplasty Registry.

Eur Heart J Qual Care Clin Outcomes. 2016;2:91-98.

II. Edfors R, Sahlen A, Szummer K, Renlund H, Evans M, Carrero JJ, Spaak J, James SK, Lagerqvist B, Varenhorst C and Jernberg T. Outcomes in patients treated with ticagrelor versus clopidogrel after acute myocardial infarction stratified by renal function.

Heart. 2018;104:1575-1582.

III. Edfors R, James SK, Szummer K, Evans M, Carrero JJ, Faxén J, Persson J, Spaak J, Varenhorst C, Jernberg T, Lagerqvist B. SWEDEHEART-1-year data show no benefit of newer generation drug-eluting stents over bare-metal stents in patients with severe kidney dysfunction following percutaneous coronary intervention.

Coron Artery Dis. In press.

IV. Edfors R, Spaak J, Lindhagen L, Andell P, Baron B, Evans M, Erlinge D, James SK, Kahan T, Lindahl B, Rezeli M, Marko-Varga G, Mörtberg J, Salzinger B, Szummer K, Wallén H, Jernberg T. Use of proteomics to identify biomarkers associated with chronic kidney disease and long-term outcomes in patients with myocardial infarction.

Manuscript.

CONTENTS

1 INTRODUCTION ... 9

1.1 Background ... 9

1.2 Definition and staging of kidney disease ... 9

1.3 Assessment of kidney function ... 10

1.4 Prognosis of RD patients in the general- and CVD population ... 12

1.5 Prognosis following CABG ... 12

1.6 Prognosis after PCI ... 13

1.7 PCI versus CABG in patients with RD... 13

1.8 Risk factors and pathological mechanisms in patients with RD and CVD .... 14

1.9 Coronary anatomy and calcification ... 14

1.10 Traditional and non-traditional risk factors ... 15

1.11 Management of CAD in patients with RD ... 16

1.12 Antithrombotic therapy ... 16

1.13 Clopidogrel... 17

1.14 Prasugrel ... 17

1.15 Ticagrelor ... 18

1.16 Interventional treatment with PCI ... 19

1.17 Bare metal stents ... 19

1.18 Older generation drug-eluting stents ... 19

1.19 Newer generation drug-eluting stents ... 20

1.20 Challenges in RD and CAD ... 21

2 AIMS... 22

3 THESIS AT A GLANCE ... 23

4 METHODS ... 24

4.1 Data sources ... 24

4.2 SWEDEHEART ... 24

4.3 Proteomics and biomarkers ... 25

4.4 Proximity extension assay... 25

4.5 Multiple reaction monitoring assay ... 26

4.6 Definitions ... 26

4.7 Definitions of outcomes ... 26

4.8 Definitions of coronary stent types... 27

4.9 Definitions of renal dysfunction ... 27

4.10 Study population... 28

4.11 Study I ... 28

4.12 Study II ... 28

4.13 Study III ... 29

4.14 Study IV ... 29

4.15 Statistics ... 29

4.16 Study I ... 29

4.17 Study II ... 30

4.18 Study III ... 30

4.19 Study IV ... 31

4.20 Ethical considerations ... 31

5 RESULTS... 49

5.1 Study I ... 49

5.2 Study II ... 52

5.3 Study III ... 53

5.4 Study IV ... 55

CONTENTS

1.1 Background ... 9

1.2 Definition and staging of kidney disease ... 9

1.3 Assessment of kidney function ... 10

1.4 Prognosis of RD patients in the general- and CVD population ... 12

1.5 Prognosis following CABG ... 12

1.6 Prognosis after PCI ... 13

1.7 PCI versus CABG in patients with RD... 13

1.8 Risk factors and pathological mechanisms in patients with RD and CVD .... 14

1.9 Coronary anatomy and calcification ... 14

1.10 Traditional and non-traditional risk factors ... 15

1.11 Management of CAD in patients with RD ... 16

1.12 Antithrombotic therapy ... 16

1.13 Clopidogrel... 17

1.14 Prasugrel ... 17

1.15 Ticagrelor ... 18

1.16 Interventional treatment with PCI ... 19

1.17 Bare metal stents ... 19

1.18 Older generation drug-eluting stents ... 19

1.19 Newer generation drug-eluting stents ... 20

1.20 Challenges in RD and CAD ... 21

2 AIMS... 22

3 THESIS AT A GLANCE ... 23

4 METHODS ... 24

4.1 Data sources ... 24

4.2 SWEDEHEART ... 24

4.3 Proteomics and biomarkers ... 25

4.4 Proximity extension assay... 25

4.5 Multiple reaction monitoring assay ... 26

4.6 Definitions ... 26

4.7 Definitions of outcomes ... 26

4.8 Definitions of coronary stent types... 27

4.9 Definitions of renal dysfunction ... 27

4.10 Study population... 28

4.11 Study I ... 28

4.12 Study II ... 28

4.13 Study III ... 29

4.14 Study IV ... 29

4.15 Statistics ... 29

4.16 Study I ... 29

4.17 Study II ... 30

4.18 Study III ... 30

4.19 Study IV ... 31

4.20 Ethical considerations ... 31

5 RESULTS... 49

5.1 Study I ... 49

5.2 Study II ... 52

5.3 Study III ... 53

5.4 Study IV ... 55

6 DISCUSSION ... 61

6.1 Major findings ... 61

6.2 Influence of RD on CV outcomes in patients with stable CAD ... 61

6.3 Management of CAD patients with concomittant RD ... 63

6.4 Dual antiplatelet treatment in relation to RD ... 63

6.5 Stent choice in CAD patients in relation to RD... 64

6.6 Proteomics to predict adverse outcomes in MI ... 65

6.7 Limitations... 67

7 CONCLUSION ... 68

8 TACK ... 69

9 REFERENCES... 71

6 DISCUSSION ... 61

6.1 Major findings ... 61

6.2 Influence of RD on CV outcomes in patients with stable CAD ... 61

6.3 Management of CAD patients with concomittant RD ... 63

6.4 Dual antiplatelet treatment in relation to RD ... 63

6.5 Stent choice in CAD patients in relation to RD... 64

6.6 Proteomics to predict adverse outcomes in MI ... 65

6.7 Limitations... 67

7 CONCLUSION ... 68

8 TACK ... 69

9 REFERENCES... 71

LIST OF ABBREVIATIONS

ACEi Angiotensin-converting enzyme inhibitor ACR Albumin creatinine ratio

ACS Acute coronary syndrome

ADP Adenosine diphosphate

ARB Angiotensin II receptor blocker

ARC Academic Research Consortium

ASA Acetylsalicylic acid

BMS Bare metal stent

CABG Coronary artery bypass surgery

CAD Coronary artery disease

CCU Coronary care unit

CG Cockcroft and Gault formula

CI Confidence interval

CKD Chronic kidney disease

CKD-EPI Chronic Kidney Disease Epidemiology Collaboration formula

CV Cardiovascular

DES Drug eluting stent

DM Diabetes Mellitus

eGFR Estimated glomerular filtration rate ESC European Society of Cardiology ESRD End stage renal disease

FABP Fatty acid binding protein 4 (adipocyte) FGF-23 Fibroblast growth factor 23

GDF-15 Growth differentiation factor 15

HD Hemodialysis

HDL High-density lipoprotein

HR Hazard ratio

HRPR High residual platelet reactivity

HT Hypertension

IHD Ischemic heart disease

IQR Interquartile range

KDIGO Kidney Disease: Improving Global Outcomes

LIST OF ABBREVIATIONS

ACEi Angiotensin-converting enzyme inhibitor ACR Albumin creatinine ratio

ACS Acute coronary syndrome

ADP Adenosine diphosphate

ARB Angiotensin II receptor blocker

ARC Academic Research Consortium

ASA Acetylsalicylic acid

BMS Bare metal stent

CABG Coronary artery bypass surgery

CAD Coronary artery disease

CCU Coronary care unit

CG Cockcroft and Gault formula

CI Confidence interval

CKD Chronic kidney disease

CKD-EPI Chronic Kidney Disease Epidemiology Collaboration formula

CV Cardiovascular

DES Drug eluting stent

DM Diabetes Mellitus

eGFR Estimated glomerular filtration rate ESC European Society of Cardiology ESRD End stage renal disease

FABP Fatty acid binding protein 4 (adipocyte) FGF-23 Fibroblast growth factor 23

GDF-15 Growth differentiation factor 15

HD Hemodialysis

HDL High-density lipoprotein

HR Hazard ratio

HRPR High residual platelet reactivity

HT Hypertension

IHD Ischemic heart disease

IQR Interquartile range

KDIGO Kidney Disease: Improving Global Outcomes

LDL Low-density lipoprotein LVEF Left ventricular ejection fraction

MDRD Modification of Diet in Renal Disease (MDRD) formula

MI Myocardial infarction

MRM Multiple Reaction Monitoring

N Numbers

N-DES New generation drug eluting stent NSTEMI Non-ST elevation myocardial infarction

NO Nitric oxide

NYHA New York Heart Association O-DES Old generation drug eluting stent

OR Odds ratio

P2Y12 P2Y purinoceptor 12

PCI Percutaneous coronary intervention PCR Polymerase chain reaction

PEA Proximity Extension Assay

PTCA Balloon angioplasty

RCT Randomized controlled trial

RD Renal dysfunction

RRT Renal replacement therapy

SCAAR Swedish Coronary Angiography and Angioplasty Registry SCAD Stable coronary artery disease

SIS Stable isotope standards

SWEDEHEART Swedish Web-System for Enhancement and Development of Evidence-Based Care in Heart Disease Evaluated According to Recommended Therapies

ST Stent thrombosis

STEMI ST-elevation myocardial infarction

UA Unstable angina pectoris

SIS Stable isotope standards

SD Standard deviation

TNF Tumor necrosis factor

TNF-R1 Tumor necrosis factor receptor 1 TNF-R2 Tumor necrosis factor receptor 2

TRAIL-2 Tumor necrosis factor related apoptosis-inducing ligand receptor 2

LDL Low-density lipoprotein

LVEF Left ventricular ejection fraction

MDRD Modification of Diet in Renal Disease (MDRD) formula

MI Myocardial infarction

MRM Multiple Reaction Monitoring

N Numbers

N-DES New generation drug eluting stent NSTEMI Non-ST elevation myocardial infarction

NO Nitric oxide

NYHA New York Heart Association O-DES Old generation drug eluting stent

OR Odds ratio

P2Y12 P2Y purinoceptor 12

PCI Percutaneous coronary intervention PCR Polymerase chain reaction

PEA Proximity Extension Assay

PTCA Balloon angioplasty

RCT Randomized controlled trial

RD Renal dysfunction

RRT Renal replacement therapy

SCAAR Swedish Coronary Angiography and Angioplasty Registry SCAD Stable coronary artery disease

SIS Stable isotope standards

SWEDEHEART Swedish Web-System for Enhancement and Development of Evidence-Based Care in Heart Disease Evaluated According to Recommended Therapies

ST Stent thrombosis

STEMI ST-elevation myocardial infarction

UA Unstable angina pectoris

SIS Stable isotope standards

SD Standard deviation

TNF Tumor necrosis factor

TNF-R1 Tumor necrosis factor receptor 1 TNF-R2 Tumor necrosis factor receptor 2

TRAIL-2 Tumor necrosis factor related apoptosis-inducing ligand receptor 2

1 INTRODUCTION

1.1 BACKGROUND

Renal dysfunction (RD) is a major public health problem with a prevalence of 10-15% in the general population of developed countries ¹. Ischemic heart disease (IHD) is the leading cause of death in patients with RD and cardiovascular (CV) mortality is greater than mortality due to the progression of end-stage renal insufficiency ². RD is also common in patients admitted to hospital for acute coronary syndrome ³, where about one third have reduced renal function.

The early stages of RD are usually asymptomatic and symptoms may often first appear together with complications in the later stages of the disease. Due to an ageing population with associated comorbidities and improved treatment regimens of ACS, the prevalence of patients with concomitant RD and coronary artery disease (CAD) is increasing ⁴⁵. Patients with RD and concomitant IHD not only have a poor prognosis, they also suffer from increased risk of bleeding, systemic drug toxicity, infection and adverse effects of interventions used to prevent or treat the condition ^6-10.

Very few randomized cardiovascular trials have included patients with RD and many recommendations concerning patients with RD and IHD are based on extrapolation of data from the general population ^{11, 12}, leading to a gap in evidence-based treatment regimens for these patients. This has important treatment implications, as measures directed at preventing the progression of RD may also prevent cardiovascular morbidity and mortality. Due to a lack of knowledge of how to treat patients with concomitant RD and IHD, further evaluations of therapies in clinical practice and associated outcomes from real-world registries are important.

1.2 DEFINITION AND STAGING OF KIDNEY DISEASE

Chronic kidney disease (CKD) is defined as pathology of the kidney structure or altered kidney function for a period of more than 3 months. According to the Kidney Disease:

Improving Global Outcomes (KDIGO) ¹³ guidelines, the definition of CKD includes patients with evidence of kidney damage (albuminuria with an albumin:creatinine ratio (ACR) > 3 mg/mmol), hematuria, tubular dysfunction associated electrolyte abnormalities, histological abnormalities, structural damage detected by imaging modalities, prior kidney transplantation or patients with glomerular filtration rate (GFR) of less than 60 ml/min/1.73m² on at least two occasions 90 days apart. CKD is classified based on the GFR and the level of proteinuria and helps to risk-stratify patients. Patients are classified as G1- G5, based on the GFR, and A1-A3 based on the ACR (Figure 1).

1 INTRODUCTION

1.1 BACKGROUND

Renal dysfunction (RD) is a major public health problem with a prevalence of 10-15% in the general population of developed countries ¹. Ischemic heart disease (IHD) is the leading cause of death in patients with RD and cardiovascular (CV) mortality is greater than mortality due to the progression of end-stage renal insufficiency ². RD is also common in patients admitted to hospital for acute coronary syndrome ³, where about one third have reduced renal function.

The early stages of RD are usually asymptomatic and symptoms may often first appear together with complications in the later stages of the disease. Due to an ageing population with associated comorbidities and improved treatment regimens of ACS, the prevalence of patients with concomitant RD and coronary artery disease (CAD) is increasing ⁴⁵. Patients with RD and concomitant IHD not only have a poor prognosis, they also suffer from increased risk of bleeding, systemic drug toxicity, infection and adverse effects of interventions used to prevent or treat the condition ^6-10.

Very few randomized cardiovascular trials have included patients with RD and many recommendations concerning patients with RD and IHD are based on extrapolation of data from the general population ^{11, 12}, leading to a gap in evidence-based treatment regimens for these patients. This has important treatment implications, as measures directed at preventing the progression of RD may also prevent cardiovascular morbidity and mortality. Due to a lack of knowledge of how to treat patients with concomitant RD and IHD, further evaluations of therapies in clinical practice and associated outcomes from real-world registries are important.

1.2 DEFINITION AND STAGING OF KIDNEY DISEASE

Chronic kidney disease (CKD) is defined as pathology of the kidney structure or altered kidney function for a period of more than 3 months. According to the Kidney Disease:

Improving Global Outcomes (KDIGO) ¹³ guidelines, the definition of CKD includes patients with evidence of kidney damage (albuminuria with an albumin:creatinine ratio (ACR) > 3 mg/mmol), hematuria, tubular dysfunction associated electrolyte abnormalities, histological abnormalities, structural damage detected by imaging modalities, prior kidney transplantation or patients with glomerular filtration rate (GFR) of less than 60 ml/min/1.73m² on at least two occasions 90 days apart. CKD is classified based on the GFR and the level of proteinuria and helps to risk-stratify patients. Patients are classified as G1- G5, based on the GFR, and A1-A3 based on the ACR (Figure 1).

Figure 1. Classification of Chronic Kidney Disease ¹⁴ based on GFR (glomerular filtration rate) categories G1-G5 and ACR (albumin:creatinine ratio) categories A1-A3. Moderate risk (yellow) – 73% of patients with CKD, High risk (orange) – 18% of patients with CKD, Very high risk (red) – 9% of patients with CKD.

Reproduced with permission from Elsevier¹⁵.

1.3 ASSESSMENT OF KIDNEY FUNCTION

Creatinine measurement is perhaps the easiest way to assess renal function but it is often not a reliable method to use since age, muscle mass, nutrition status, obesity and hydration may affect creatinine levels. An overestimation of renal function may occur in, for example, the elderly, females or individuals with low body weight. GFR is the key indicator of renal function and the most accurate method to obtain GFR is to inject a non-metabolizable and non-absorbent exogenous markers such as inulin or ioxehol and subsequent concentration measurement in the urine or plasma. In clinical practice this may, however, often prove to be impractical, time-ineffective and in many clinical situations the precis knowledge of GFR is not always necessary.

Various mathematical equations that use serum creatinine and combinations of age, sex, race and weight have been developed to obtain estimated GFR (eGFR). A number of recognized and well-validated formulae including the Chronic Kidney Disease Epidemiology (CKD- EPI) equation¹⁴, the Modification of Diet in Renal Disease (MDRD) study equations ^{16, 17} and

Figure 1. Classification of Chronic Kidney Disease ¹⁴ based on GFR (glomerular filtration rate) categories G1-G5 and ACR (albumin:creatinine ratio) categories A1-A3. Moderate risk (yellow) – 73% of patients with CKD, High risk (orange) – 18% of patients with CKD, Very high risk (red) – 9% of patients with CKD.

Reproduced with permission from Elsevier¹⁵.

1.3 ASSESSMENT OF KIDNEY FUNCTION

Creatinine measurement is perhaps the easiest way to assess renal function but it is often not a reliable method to use since age, muscle mass, nutrition status, obesity and hydration may affect creatinine levels. An overestimation of renal function may occur in, for example, the elderly, females or individuals with low body weight. GFR is the key indicator of renal function and the most accurate method to obtain GFR is to inject a non-metabolizable and non-absorbent exogenous markers such as inulin or ioxehol and subsequent concentration measurement in the urine or plasma. In clinical practice this may, however, often prove to be impractical, time-ineffective and in many clinical situations the precis knowledge of GFR is not always necessary.

Various mathematical equations that use serum creatinine and combinations of age, sex, race and weight have been developed to obtain estimated GFR (eGFR). A number of recognized and well-validated formulae including the Chronic Kidney Disease Epidemiology (CKD- EPI) equation¹⁴, the Modification of Diet in Renal Disease (MDRD) study equations ^{16, 17} and

the Cockcroft-Gault equation (CG) ¹⁸ are often utilized. The CG equation was introduced prior to the use of standardized creatinine assays and is in contrast to MDRD and CKD-EPI, not adapted for use with creatinine values traceable to standardized reference materials. In contrast to CKD-EPI and MDRD, CG includes weight but not race and eGFR is expressed in ml/min which is not normalized to body surface area of 1.73m².

In a study of 117 healthy individuals that underwent routine kidney donor evaluation, the ability of the MDRD and the CG equation was evaluated to predict GFR assessed by (125)I- iothalamate or (99m)Tc-diethylenetriamine-pentaacetic acid (DTPA)¹⁹. The study found that the MDRD equation underestimates GFR, whereas the CG equation overestimates GFR in people with normal renal function. In order to provide a more accurate estimate of GFR in patients with normal or only mildly reduced GFR, the CKD-EPI equation was developed and was associated with less bias, improved precision, and greater accuracy compared to the MDRD study ¹⁴. In a meta-analysis that included 1 million adults from the general population as well as high risk cohorts, CKD-EPI versus MDRD equations resulted in a lower prevalence of CKD but a more accurate risk prediction for adverse outcomes was found ²⁰.

The main goal of using eGFR is to diagnose and classify CKD, to adjust dosage of medication, and to predict adverse prognosis. The Kidney Disease: Improving Global Outcomes ¹³¹³ and National Institute for Health Excellence recommend using the creatinine derived CKD-EPI equation, since it gives the most accurate renal function estimate compared to GFR ²¹. Even though in 2010 the Food and Drug Administration changed the guidance for the industry to allow MDRD estimated GFR for drug dosing²², CG estimated renal dose adjustment is widely used for the purpose of dose adjustments.

However, for prognosis, the choice of eGFR equation is less clear. In patients with heart failure, CG predicted mortality better than CKD-EPI and MDRD ²³ and in another study CG versus MDRD was superior in identifying more patients with MI at higher risk ²⁴. In another study of patients with cardio-vascular disease (CVD), CKD-EPI as compared to MDRD more often diagnosed RD, and reclassified patients to a higher risk group and more accurate risk stratification ²⁵.

Some observational studies define RD according to International Classification of Diseases (ICD) codes for renal replacement therapy initiation or other related conditions such as hospitalization for RD ^{26, 27}. The use of ICD codes instead of eGFR will most often underestimate RD prevalence ²⁸ and offers no possibility to assess risks associated with RD prior to renal replacement therapy.

In many ACS studies, the baseline creatinine value at the time of ACS presentation is often used to assess the GFR and the presence of RD. However, KDIGO ¹³ guidelines recommend creatinine to be measured when patients are in a stable condition in order to avoid the issue that kidney function instead reflects the ischemic severity of ACS. This is a limitation to observational studies in general as well as to this thesis, even though SWEDEHEART instructions for data collection recommend clinicians to use a creatinine value that best reflects the patients underlying condition. Documentation of duration is usually not available

the Cockcroft-Gault equation (CG) ¹⁸ are often utilized. The CG equation was introduced prior to the use of standardized creatinine assays and is in contrast to MDRD and CKD-EPI, not adapted for use with creatinine values traceable to standardized reference materials. In contrast to CKD-EPI and MDRD, CG includes weight but not race and eGFR is expressed in ml/min which is not normalized to body surface area of 1.73m².

In a study of 117 healthy individuals that underwent routine kidney donor evaluation, the ability of the MDRD and the CG equation was evaluated to predict GFR assessed by (125)I- iothalamate or (99m)Tc-diethylenetriamine-pentaacetic acid (DTPA)¹⁹. The study found that the MDRD equation underestimates GFR, whereas the CG equation overestimates GFR in people with normal renal function. In order to provide a more accurate estimate of GFR in patients with normal or only mildly reduced GFR, the CKD-EPI equation was developed and was associated with less bias, improved precision, and greater accuracy compared to the MDRD study ¹⁴. In a meta-analysis that included 1 million adults from the general population as well as high risk cohorts, CKD-EPI versus MDRD equations resulted in a lower prevalence of CKD but a more accurate risk prediction for adverse outcomes was found ²⁰.

The main goal of using eGFR is to diagnose and classify CKD, to adjust dosage of medication, and to predict adverse prognosis. The Kidney Disease: Improving Global Outcomes ¹³¹³ and National Institute for Health Excellence recommend using the creatinine derived CKD-EPI equation, since it gives the most accurate renal function estimate compared to GFR ²¹. Even though in 2010 the Food and Drug Administration changed the guidance for the industry to allow MDRD estimated GFR for drug dosing²², CG estimated renal dose adjustment is widely used for the purpose of dose adjustments.

However, for prognosis, the choice of eGFR equation is less clear. In patients with heart failure, CG predicted mortality better than CKD-EPI and MDRD ²³ and in another study CG versus MDRD was superior in identifying more patients with MI at higher risk ²⁴. In another study of patients with cardio-vascular disease (CVD), CKD-EPI as compared to MDRD more often diagnosed RD, and reclassified patients to a higher risk group and more accurate risk stratification ²⁵.

Some observational studies define RD according to International Classification of Diseases (ICD) codes for renal replacement therapy initiation or other related conditions such as hospitalization for RD ^{26, 27}. The use of ICD codes instead of eGFR will most often underestimate RD prevalence ²⁸ and offers no possibility to assess risks associated with RD prior to renal replacement therapy.

In many ACS studies, the baseline creatinine value at the time of ACS presentation is often used to assess the GFR and the presence of RD. However, KDIGO ¹³ guidelines recommend creatinine to be measured when patients are in a stable condition in order to avoid the issue that kidney function instead reflects the ischemic severity of ACS. This is a limitation to observational studies in general as well as to this thesis, even though SWEDEHEART instructions for data collection recommend clinicians to use a creatinine value that best reflects the patients underlying condition. Documentation of duration is usually not available

associated with increased risk of death even in the absence of albuminuria ^{28, 29}, an eGFR below 60 mL/min/1.73 m² is often defined as having CKD in routine clinical practice.

1.4 PROGNOSIS OF RD PATIENTS IN THE GENERAL- AND CVD POPULATION Renal origin of CV was first suggested by Richard Bright as early as in 1836. This has been confirmed by multiple epidemiological studies, where it is evident that patients with RD are at a high risk of developing CVD ³⁰. CVD, and more specifically IHD is a major cause of death and morbidity in patients with RD. In a population-based study of more than 1 million individuals, the adjusted Hazard Ratio (HR) for death increased with the rate of eGFR in an inverse fashion; HR 1.2, 95% confidence interval (CI) (1.1-1.2) for eGFR of 45–59 mL/min/1.73 m², HR 1.8, 95% CI (1.7-1.9) for eGFR of 30–44 mL/min/1.73 m², HR 3.2, 95% CI (3.1-3.4) for eGFR of 15–29 mL/min/1.73 m2 and HR 5.9, 95% CI (5.4-6.5) for eGFR

<15 mL/min/1.73 m², respectively and the risk of cardiovascular events followed the same trend ². In a meta-analysis of 10 cohorts including more than 200,000 patients, eGFR as well as albuminuria were closely associated with the risk of all-cause mortality ²⁹. For every 10 mL/min/1.73 m² reduction in eGFR, the risk for cardiovascular mortality is increased by 5%

31. Albuminuria has an independent and additive effect on reduced eGFR regarding the risk of cardiovascular mortality ³².

Patients with end-stage renal disease (ESRD) and the early stages RD are associated with high rates of morbidity, despite similar rates of comorbidities, which may imply that complications of ESRD are evident before the onset of ESRD ³³. According to data from the Swedish Renal Registry the overall 5-year patient survival rate for patients on renal replacement therapy (RRT) is 23.1% whereas the 5-year survival rate of after kidney transplantation is about 85% ³⁴.

There is a strong relationship between RD, CV outcomes and mortality in patients with ACS

103. The prognosis of RRT patients suffering from ACS is extremely poor, where 1-year survival rates of 41% have been described³⁵.

1.5 PROGNOSIS FOLLOWING CABG

Renal function is incorporated in the European System for Cardiac Operative Risk Evaluation (EuroSCORE) ³⁶, and several studies demonstrate that RD is associated with worse outcomes following Coronary artery bypass grafting (CABG) ^{37, 38}. Even mild renal dysfunction defined as Creatinine >130 µmol/L, is an independent risk factor for adverse outcome after CABG ³⁹. Patients with RD are at risk of developing acute kidney injury and may require RRT temporary or permanently following surgery. In particular, in patients with RD and history of diabetes, peripheral artery disease ⁴⁰ and RRT there is a high risk of in- hospital complications and mortality ⁴¹.

associated with increased risk of death even in the absence of albuminuria ^{28, 29}, an eGFR below 60 mL/min/1.73 m² is often defined as having CKD in routine clinical practice.

1.4 PROGNOSIS OF RD PATIENTS IN THE GENERAL- AND CVD POPULATION Renal origin of CV was first suggested by Richard Bright as early as in 1836. This has been confirmed by multiple epidemiological studies, where it is evident that patients with RD are at a high risk of developing CVD ³⁰. CVD, and more specifically IHD is a major cause of death and morbidity in patients with RD. In a population-based study of more than 1 million individuals, the adjusted Hazard Ratio (HR) for death increased with the rate of eGFR in an inverse fashion; HR 1.2, 95% confidence interval (CI) (1.1-1.2) for eGFR of 45–59 mL/min/1.73 m², HR 1.8, 95% CI (1.7-1.9) for eGFR of 30–44 mL/min/1.73 m², HR 3.2, 95% CI (3.1-3.4) for eGFR of 15–29 mL/min/1.73 m2 and HR 5.9, 95% CI (5.4-6.5) for eGFR

<15 mL/min/1.73 m², respectively and the risk of cardiovascular events followed the same trend ². In a meta-analysis of 10 cohorts including more than 200,000 patients, eGFR as well as albuminuria were closely associated with the risk of all-cause mortality ²⁹. For every 10 mL/min/1.73 m² reduction in eGFR, the risk for cardiovascular mortality is increased by 5%

31. Albuminuria has an independent and additive effect on reduced eGFR regarding the risk of cardiovascular mortality ³².

Patients with end-stage renal disease (ESRD) and the early stages RD are associated with high rates of morbidity, despite similar rates of comorbidities, which may imply that complications of ESRD are evident before the onset of ESRD ³³. According to data from the Swedish Renal Registry the overall 5-year patient survival rate for patients on renal replacement therapy (RRT) is 23.1% whereas the 5-year survival rate of after kidney transplantation is about 85% ³⁴.

There is a strong relationship between RD, CV outcomes and mortality in patients with ACS

103. The prognosis of RRT patients suffering from ACS is extremely poor, where 1-year survival rates of 41% have been described³⁵.

1.5 PROGNOSIS FOLLOWING CABG

Renal function is incorporated in the European System for Cardiac Operative Risk Evaluation (EuroSCORE) ³⁶, and several studies demonstrate that RD is associated with worse outcomes following Coronary artery bypass grafting (CABG) ^{37, 38}. Even mild renal dysfunction defined as Creatinine >130 µmol/L, is an independent risk factor for adverse outcome after CABG ³⁹. Patients with RD are at risk of developing acute kidney injury and may require RRT temporary or permanently following surgery. In particular, in patients with RD and history of diabetes, peripheral artery disease ⁴⁰ and RRT there is a high risk of in- hospital complications and mortality ⁴¹.

1.6 PROGNOSIS AFTER PCI

The risk of complications, morbidity and mortality following percutaneous coronary intervention (PCI) is increased in CAD patients with RD as compared to patients with normal renal function ^{10, 42}. Even mild RD defined as Creatinine 97-130 µmol/L predicts adverse outcome following PCI ⁴³. The use of contrast media during PCI is a common cause of acute kidney injury and patients with RD often have comorbidities that may increase the risk of periprocedural ischemic and bleeding events. Since RD defines a high risk population PCI is frequently underused in patients with renal dysfunction ², even though data support that patients with renal RD and multivessel disease that undergo PCI have better survival compared to medical therapy ⁴⁴.

1.7 PCI VERSUS CABG IN PATIENTS WITH RD

Patients with RD are often excluded from randomized controlled trials (RCT) on myocardial revascularization and current data are mostly based on observational studies and post-hoc analyses from randomized trials. Few studies have specifically compared outcomes with PCI versus CABG in patients with moderate to severe RD.

CABG versus PCI in patients with moderate RD are associated with an increased risk of perioperative and short-term mortality (1 year), but lower medium-to-long-term mortality after CABG compared with PCI ⁴⁵⁴⁶. However, observational 5-year data from the United States Renal Data System of 21,981 patients with ESRD and multivessel coronary disease showed that CABG versus PCI was associated with significantly lower risks for death (HR 0.87, 95% CI (0.84-0.90)) and the composite of death or MI (HR 0.88, 95% CI (0.86-0.91))

47. In the randomized Arterial Revascularization Therapies Study ⁴⁸ post hoc analysis of patients with RD (25% of 1,205 patients) defined as estimated creatinine clearance

≤60 mL/min, CABG and multivessel PCI with bare metal stents (BMS) in patients with stable angina were compared. At 3 years of follow-up, no difference for the primary endpoint of death, MI or stroke (19% vs. 17%; HR 0.93, 95% CI (0.54–1.61)) was observed, but CABG versus PCI showed a lower risk of repeat revascularization (25% vs. 8%; HR 0.28, 95% CI (0.14–0.54))⁴⁸. Similar results were also observed at 5 years follow-up, where higher rates of repeat revascularization with PCI compared to CABG were noted ⁴⁹. In the non-randomized ARTS II study ⁵⁰, drug eluting stents (DES) with sirolimus were compared with BMS and CABG. The 5-year results showed that the event-free survival rate of the combined outcome (Death, MI or stroke) was 87.1% in DES group, versus 86.0% (p = 0.1) and 81.9% (p = 0.007) in the CABG and BMS cohorts, respectively. However, data on patients with RD were not available.

In the CKD substudy of Synergy between Percutaneous Coronary Intervention with Taxus and Cardiac Surgery (SYNTAX) ⁵¹, the 5-year results of patients randomized to either sirolimus DES or CABG were studied and compared to CABG, PCI

showed significantly higher rates of the combined endpoint (42.1% vs. 31.5%, p=0.019), mainly driven by repeat revascularization (21.9% vs. 8.9%, p=0.004) and all-cause mortality (26.7% vs. 21.2%, p=0.14). The results of the aforementioned studies may be less applicable

1.6 PROGNOSIS AFTER PCI

The risk of complications, morbidity and mortality following percutaneous coronary intervention (PCI) is increased in CAD patients with RD as compared to patients with normal renal function ^{10, 42}. Even mild RD defined as Creatinine 97-130 µmol/L predicts adverse outcome following PCI ⁴³. The use of contrast media during PCI is a common cause of acute kidney injury and patients with RD often have comorbidities that may increase the risk of periprocedural ischemic and bleeding events. Since RD defines a high risk population PCI is frequently underused in patients with renal dysfunction ², even though data support that patients with renal RD and multivessel disease that undergo PCI have better survival compared to medical therapy ⁴⁴.

1.7 PCI VERSUS CABG IN PATIENTS WITH RD

Patients with RD are often excluded from randomized controlled trials (RCT) on myocardial revascularization and current data are mostly based on observational studies and post-hoc analyses from randomized trials. Few studies have specifically compared outcomes with PCI versus CABG in patients with moderate to severe RD.

CABG versus PCI in patients with moderate RD are associated with an increased risk of perioperative and short-term mortality (1 year), but lower medium-to-long-term mortality after CABG compared with PCI ⁴⁵⁴⁶. However, observational 5-year data from the United States Renal Data System of 21,981 patients with ESRD and multivessel coronary disease showed that CABG versus PCI was associated with significantly lower risks for death (HR 0.87, 95% CI (0.84-0.90)) and the composite of death or MI (HR 0.88, 95% CI (0.86-0.91))

47. In the randomized Arterial Revascularization Therapies Study ⁴⁸ post hoc analysis of patients with RD (25% of 1,205 patients) defined as estimated creatinine clearance

≤60 mL/min, CABG and multivessel PCI with bare metal stents (BMS) in patients with stable angina were compared. At 3 years of follow-up, no difference for the primary endpoint of death, MI or stroke (19% vs. 17%; HR 0.93, 95% CI (0.54–1.61)) was observed, but CABG versus PCI showed a lower risk of repeat revascularization (25% vs. 8%; HR 0.28, 95% CI (0.14–0.54))⁴⁸. Similar results were also observed at 5 years follow-up, where higher rates of repeat revascularization with PCI compared to CABG were noted ⁴⁹. In the non-randomized ARTS II study ⁵⁰, drug eluting stents (DES) with sirolimus were compared with BMS and CABG. The 5-year results showed that the event-free survival rate of the combined outcome (Death, MI or stroke) was 87.1% in DES group, versus 86.0% (p = 0.1) and 81.9% (p = 0.007) in the CABG and BMS cohorts, respectively. However, data on patients with RD were not available.

In the CKD substudy of Synergy between Percutaneous Coronary Intervention with Taxus and Cardiac Surgery (SYNTAX) ⁵¹, the 5-year results of patients randomized to either sirolimus DES or CABG were studied and compared to CABG, PCI

showed significantly higher rates of the combined endpoint (42.1% vs. 31.5%, p=0.019), mainly driven by repeat revascularization (21.9% vs. 8.9%, p=0.004) and all-cause mortality (26.7% vs. 21.2%, p=0.14). The results of the aforementioned studies may be less applicable

in current clinical practice due to the emerging use of the newer generation of DES. Results from the randomized Bypass Surgery Versus Everolimus-Eluting Stent Implantation Multivessel Coronary Artery Disease (BEST) study ⁵² showed that PCI with the newer generation of DES (everolimus) versus CABG, was associated with an increased risk of repeat revascularization and spontaneous myocardial infarction but without any difference in death, however RD was not considered in the study. In a retrospective propensity-score matched study of 5,920 patients with eGFR<60, CABG was associated with higher short- term risk of death, stroke, and repeat revascularization, whereas PCI with everolimus-eluting stents showed higher risk of long-term risk of repeat revascularization MI but not death ⁵³. Moreover, the CKD subgroup analysis of the multicenter randomized Evaluation of XIENCE Versus Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization trial ⁵⁴ (EXCEL), showed that there were no significant differences in the rates of the composite outcome (death, MI or stroke) after PCI with the newer generation of DES versus CABG, 23.4% versus 18.1% respectively; HR 1.25; 95% CI (0.79 to 1.98).

Before the results from the ongoing RCT on optimal long-term revascularization strategies in patients with stress-induced ischemia in the International Study of Comparative Health Effectiveness With Medical and Invasive Approaches—Chronic Kidney Disease ^{55 56}, the general condition, life expectancy, the least invasive appropriate method and frailty must be considered when selecting the most appropriate revascularization strategy.

1.8 RISK FACTORS AND PATHOLOGICAL MECHANISMS IN PATIENTS WITH RD AND CVD

The mechanisms involved in the interplay between RD and CAD are numerous and not yet fully understood. One possible explanation is that patients with RD that experience MI, are more likely to have co-existence of other predictors of adverse outcomes like frailty factors, age, hypotension, and lower body weight. These factors may serve as residual confounding in various studies despite advanced statistical adjustments. Moreover, it is possible that an adverse outcome such as death or cardiovascular mortality in the presence of RD may instead serve as an indicator of a more severe ACS ⁵⁷. However, there are numerous risk factors associated with having RD.

1.9 CORONARY ANATOMY AND CALCIFICATION

It has been postulated that RD accelerates atherosclerosis and coronary artery calcification through processes associated with both traditional CVD risk factors (diabetes, hypertension, dyslipidemia, smoking status, sex and prior history of MI) and non-traditional risk factors for CVD, such as bone and mineral disorders as well as systemic inflammation¹². RD is associated with increased inflammatory activity, which may increase the risk of plaque rupture and a more pro-thrombotic state, which may increase the risk of MI when a plaque rupture occurs ^{58, 59}.

in current clinical practice due to the emerging use of the newer generation of DES. Results from the randomized Bypass Surgery Versus Everolimus-Eluting Stent Implantation Multivessel Coronary Artery Disease (BEST) study ⁵² showed that PCI with the newer generation of DES (everolimus) versus CABG, was associated with an increased risk of repeat revascularization and spontaneous myocardial infarction but without any difference in death, however RD was not considered in the study. In a retrospective propensity-score matched study of 5,920 patients with eGFR<60, CABG was associated with higher short- term risk of death, stroke, and repeat revascularization, whereas PCI with everolimus-eluting stents showed higher risk of long-term risk of repeat revascularization MI but not death ⁵³. Moreover, the CKD subgroup analysis of the multicenter randomized Evaluation of XIENCE Versus Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization trial ⁵⁴ (EXCEL), showed that there were no significant differences in the rates of the composite outcome (death, MI or stroke) after PCI with the newer generation of DES versus CABG, 23.4% versus 18.1% respectively; HR 1.25; 95% CI (0.79 to 1.98).

Before the results from the ongoing RCT on optimal long-term revascularization strategies in patients with stress-induced ischemia in the International Study of Comparative Health Effectiveness With Medical and Invasive Approaches—Chronic Kidney Disease ^{55 56}, the general condition, life expectancy, the least invasive appropriate method and frailty must be considered when selecting the most appropriate revascularization strategy.

1.8 RISK FACTORS AND PATHOLOGICAL MECHANISMS IN PATIENTS WITH RD AND CVD

The mechanisms involved in the interplay between RD and CAD are numerous and not yet fully understood. One possible explanation is that patients with RD that experience MI, are more likely to have co-existence of other predictors of adverse outcomes like frailty factors, age, hypotension, and lower body weight. These factors may serve as residual confounding in various studies despite advanced statistical adjustments. Moreover, it is possible that an adverse outcome such as death or cardiovascular mortality in the presence of RD may instead serve as an indicator of a more severe ACS ⁵⁷. However, there are numerous risk factors associated with having RD.

1.9 CORONARY ANATOMY AND CALCIFICATION

It has been postulated that RD accelerates atherosclerosis and coronary artery calcification through processes associated with both traditional CVD risk factors (diabetes, hypertension, dyslipidemia, smoking status, sex and prior history of MI) and non-traditional risk factors for CVD, such as bone and mineral disorders as well as systemic inflammation¹². RD is associated with increased inflammatory activity, which may increase the risk of plaque rupture and a more pro-thrombotic state, which may increase the risk of MI when a plaque rupture occurs ^{58, 59}.